Skip to main content
. Author manuscript; available in PMC: 2020 May 21.
Published in final edited form as: J Am Coll Cardiol. 2019 May 21;73(19):2388–2397. doi: 10.1016/j.jacc.2019.02.057

Figure 2. Calibration Plots in the Jackson Heart Study and PREVEND Validation Samples with Discrimination and Calibration Statistics.

Figure 2.

Calibration of the Pooled Cohort equations to Prevent Heart Failure (PCP-HF) risk score in 2 large population-based cohorts of White (PREVEND) and Black (Jackson Heart Study) adults aged 30-80 years after excluding individuals with prevalent CVD. The remaining participants were grouped into deciles (collapsed when fewer than 2 events were observed in any group) according to increasing predicted risk and average predicted risk is compared with average observed risk in each sex-race group. The model demonstrated good discrimination with c-statistics ranging from 0.74-0.79 and strong calibration (chi-sq <20 with p>0.05 for all sex-race groups except white men p=0.03).